A method and system for identifying faults in a heating, ventilation, and air conditioning (HVAC) device

Abstract

The application discloses a kind of heating equipment fault identification method and system, it is related to heating equipment technical field, building use behavior, environment response data and heating equipment operation data are collected, the coupling model between building behavior and environment, equipment response and behavior-driven operation mode are constructed, inconsistent area and abnormal behavior trajectory of behavior response are identified, and structured exception coding is generated, abnormal trajectory and equipment maintenance data are associated, potential fault type is identified and fault information is output, the building behavior of the present application can be modeled and analyzed comprehensively, environmental response and equipment operation data, improve the accuracy and real-time of fault identification, can accurately identify abnormal behavior trajectory and response misplacement, assist in locating potential fault cause of equipment, through structured early warning and visual display, realize the rapid feedback and operation and maintenance response of fault information, introduce causal association analysis, effectively utilize historical maintenance data, improve the intelligentization and maintenance efficiency of fault judgment.

Description

Technical Field

[0001] This invention relates to the field of HVAC equipment technology, and in particular to a method and system for identifying faults in HVAC equipment. Background Technology

[0002] Heating, ventilation, and air conditioning (HVAC) equipment is widely used in modern buildings to ensure indoor comfort and safety. However, due to the variety of equipment and the complexity of the working environment, traditional methods of equipment fault identification often rely on regular maintenance and manual judgment, which cannot detect potential equipment faults in a timely and accurate manner. This leads to excessive downtime, increased maintenance costs, and energy waste. With the development of big data and Internet of Things (IoT) technologies, intelligent analysis combining building usage behavior, environmental data, and equipment operation information has become a key technology for improving the efficiency and accuracy of HVAC equipment fault diagnosis.

[0003] Currently, Chinese invention patent application number 202110192487.8 discloses a method, apparatus, device, and storage medium for fault identification of HVAC equipment. This method utilizes a pre-set fuzzy search library and an exception jump library in the processing device. After obtaining the test item name, a judgment character is selected from the fuzzy search library and compared with the test item name. The judgment character that matches the comparison result is used as the preliminary identification result. Then, it is determined whether the exception character corresponding to the preliminary identification result in the exception jump library matches the test item name. If they match, a judgment character identical to the matching exception character is selected from the fuzzy search library as the final identification result; otherwise, the preliminary identification result is used as the final identification result. This setup, using a processing device for fault identification, can effectively improve processing efficiency compared to manual identification. Furthermore, by combining fuzzy search and exception jump functions, unique identification of HVAC equipment test item names can be achieved, improving the accuracy of the identification results.

[0004] The aforementioned technologies struggle to accurately identify response anomalies and potential faults by linking building behavior and HVAC equipment operating status in real time using multi-source data. Summary of the Invention

[0005] The technical problem solved by this invention is that existing technologies are unable to accurately identify response anomalies and potential faults by linking building behavior and HVAC equipment operation status in real time through multi-source data.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A method for identifying faults in HVAC equipment includes the following steps:

[0008] Step S1: Collect building usage behavior, environmental response data, and HVAC equipment operation data;

[0009] Step S2: Construct a coupling model and behavior-driven operation mode between building behavior and environment and equipment response;

[0010] Step S3: Identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and generate structured anomaly codes;

[0011] Step S4: Associate the abnormal trajectory with equipment maintenance data, identify potential fault types, and output fault information;

[0012] Step S5: Generate structured early warning data, visualize and track operation and maintenance responses;

[0013] Step S2 includes the following sub-steps:

[0014] Step S201: Establish a temporal correspondence between building usage behavior data and environmental response data. The logic for establishing the temporal correspondence is as follows:

[0015] The building usage behavior data and environmental response data are time-aligned, the building usage behavior data is converted into a standardized time series feature vector, and a sliding window comparison process is performed with minutes as the time window unit, comparing it with the environmental response data within the same time window.

[0016] Calculate the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data within each time window. Record the trend changes of the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data over a continuous period of time. Based on the correlation coefficient, cross-lag time difference, co-fluctuation amplitude, and trend changes between the rate of change of building use behavior data and the rate of change of environmental response data, establish a dynamic coupling model between building use behavior data and environmental response data, and extract the driving features of the dynamic coupling model. The driving features include the maximum correlation time lag, the difference in fluctuation amplitude, and the peak response duration.

[0017] Step S202: Establish a regulation response mapping between environmental response data and HVAC equipment operation data. The logic for establishing the regulation response mapping is as follows:

[0018] The environmental response data is converted into a standardized time series and time-aligned with the HVAC equipment operation data. Using minutes as the time window unit, a sliding window comparison is performed to analyze the correspondence between the parameter fluctuations of the environmental response data and the data state changes of the HVAC equipment operation data. The equipment response delay, adjustment amplitude, and response rate within each time window are calculated, and response features are extracted. The response features include adjustment lag time, adjustment intensity change amplitude, and continuous adjustment time.

[0019] Step S203: Integrate building usage behavior data, environmental response data, and equipment operation data, and use driving features and response features as joint inputs to construct a building behavior-driven operation mode. The construction logic of the building behavior-driven operation mode is as follows:

[0020] Using a sliding time window as the analysis unit, the collaborative change segments of building usage behavior data, environmental response data and equipment operation data on the time axis are identified respectively. A multivariate dynamic regression model and an adaptive weight fusion mechanism are constructed to establish a behavior-driven equipment response prediction sub-model.

[0021] By using response features as mediating variables, path analysis is introduced to quantify the moderating effect of response features on equipment operation response driven by changes in building behavior. The moderating effect index includes the direction of adjustment, intensity, and hysteresis amplitude. A linkage map is constructed, and the path intensity and response probability are dynamically updated and visualized using a data-driven approach, outputting a structured building behavior response feature vector.

[0022] Preferably, step S1 includes the following sub-steps:

[0023] Step S101: Collect building usage behavior data, which includes the frequency of personnel movement, the density of personnel movement, the passage time, the number of people passing through, the on / off status of the lighting system, and the duration of the lighting system being on.

[0024] Step S102: Collect environmental response data, including ambient temperature, ambient relative humidity, and ambient carbon dioxide concentration;

[0025] Step S103: Collect HVAC equipment operation data, which includes equipment start-up and shutdown time, equipment start-up and shutdown frequency, fan operating speed, temperature difference between supply air outlet and return air outlet, and energy consumption data.

[0026] Preferably, step S3 includes the following sub-steps:

[0027] Step S301 involves matching and analyzing the structured building behavior response feature vectors between building behavior and equipment operation. Based on the constructed dynamic mapping model, a one-to-one correspondence analysis is performed between the structured building behavior response feature vectors and the equipment operation state sequence within a unit time window. The synchronization rate is used to evaluate the temporal consistency.

[0028] If a misalignment, decrease in synchronization rate, or change in correlation coefficient from positive to negative is detected in the structured building behavior response feature vector within a continuous time window, it is considered an area of ​​inconsistent behavior response, and the corresponding spatial location identifier, start time, and duration are recorded.

[0029] Preferably, step S3 further includes the following sub-steps:

[0030] Step S302: Identify abnormal behavioral trajectories of equipment energy consumption curve deviation and response lag under specific behavioral patterns. Perform cluster analysis based on the structured building behavior response feature vector, output behavioral pattern clustering results, and select typical behavioral cycles based on the behavioral pattern clustering results. The typical behavioral cycles include high-density office work, nighttime idleness, and concentrated meetings. Perform statistical analysis on the equipment energy consumption curve and operating parameters within each typical behavioral cycle. The logic of the statistical analysis is as follows:

[0031] Compare the equipment energy consumption curve and operating parameters with the baseline operating template. If the equipment energy consumption curve shows a sudden increase, continuous high fluctuation, or significant response delay under the condition that the specific behavioral input remains unchanged, it is judged as an abnormal trajectory.

[0032] By setting preset energy consumption offset threshold and response time upper limit, the start and end time, behavior pattern label and fluctuation amplitude of abnormal trajectory are calibrated, and abnormal behavior trajectory is output.

[0033] Step S303: Jointly structure and encode the inconsistent response area and the abnormal behavior trajectory. The encoded fields of the structured encoding include the abnormality type, the scope of impact, the associated device number, the start and end timestamps, and the abnormality score.

[0034] Preferably, step S4 includes the following sub-steps:

[0035] Step S401: Obtain historical maintenance data and key component operating status data of the equipment, and retrieve maintenance records from the equipment management system. The maintenance records include key component replacement time, maintenance frequency, maintenance item details, fault description, handling method and actual handling time.

[0036] The system synchronously collects operational status data of key components, including fans, electric valves, sensors, and control execution units. The operational status data includes cumulative running time, temperature rise curve slope, vibration spectrum peak value, and abnormal fluctuation amplitude of operating current.

[0037] Step S402: Perform causal correlation analysis between the abnormal behavior trajectory and historical maintenance data. Using the associated device number and abnormal time period in the abnormal behavior trajectory output in step S303 as anchor points, retrieve the maintenance records and status fluctuation information of the corresponding device in the previous and subsequent cycles in the maintenance database.

[0038] A causal mapping rule table is constructed based on the equipment type and the corresponding component characteristics. It is determined whether there are any abnormal situations before the occurrence of abnormal trajectories. The abnormal situations include maintenance delay records, frequent failures, or significant deterioration trends in operating status indicators. Causal association is determined based on the abnormal situations, and the causal association is output. The causal association includes abnormal behavior and component aging, lack of lubrication, and transmission imbalance.

[0039] Step S403: Based on abnormal conditions, energy consumption curve deviations and changes in the status of key components, determine whether there are hidden faults in the equipment. The hidden faults include air duct blockage, fan aging, sensor accuracy degradation and control logic delay.

[0040] The final fault type identification result is output by combining the causal mapping rule table, along with equipment identification, fault classification code, impact level score and suggested handling measures.

[0041] Preferably, step S5 includes the following sub-steps:

[0042] Step S501: Convert the final fault type identification result into structured early warning data, specifically including:

[0043] Read the device identifier, fault type, risk level and recommended measures, and combine them with the spatial location coding information in the abnormal behavior trajectory to generate a standard format early warning data structure. The early warning data structure includes spatial location information, abnormal situation, hidden fault, time period and device coverage area.

[0044] Step S502: Visualize the structured early warning data and synchronize it to the operation and maintenance platform. Use layer overlay to highlight the identified abnormal areas on the building floor plan and display information cards in the corresponding areas. The information cards include the equipment number, abnormal situation, hidden fault, risk level score and recommended handling solution.

[0045] Step S503: Update the dynamic mapping model based on the actual response situation and track the operation and maintenance response records after the early warning information is released. The actual response situation includes the response start time, handling method, handling duration and final fault confirmation result.

[0046] The actual response is compared with the structured early warning data to determine whether there are false alarms, missed alarms, and classification biases, and to correct the judgment thresholds and causal reasoning path parameters.

[0047] Preferably, the causal association analysis specifically includes:

[0048] Based on the device status record data, compare the time when the adjustment signal is issued with the actual response time of the device, calculate the average response time, response variance and maximum delay time, and determine whether it exceeds the preset threshold.

[0049] The number of times the equipment starts and stops within the time period of abnormal behavior trajectory is statistically analyzed, and combined with building usage behavior data, it is determined whether there are invalid responses with frequent starts and stops but no effective environmental regulation effect.

[0050] Pattern recognition is performed on the energy consumption curves within the specified time period. The current energy consumption curve is normalized and aligned with the historical benchmark energy consumption curves under similar operating conditions, and the deviation index is calculated. Combined with equipment aging label data, a weighted comprehensive scoring model is used to calculate the risk level and output the judgment result.

[0051] Preferably, the abnormal behavior trajectory includes one or more abnormal manifestations, which are identified and confirmed by combining the time period of occurrence, corresponding behavior pattern, and operating status:

[0052] During periods of stable personnel flow and limited changes in heat load, if the equipment energy consumption data curve shows non-periodic fluctuations and a sudden increase in energy consumption without obvious external driving conditions, the equipment is judged to be ineffective in operation and energy consumption regulation failure.

[0053] The equipment frequently starts and stops without corresponding changes in building usage data, indicating that the problem stems from an abnormal start / stop logic combined with excessively high control sensitivity.

[0054] When environmental response data fluctuates rapidly, if the operating status data fails to respond in a timely manner or the delay exceeds the preset adjustment response threshold, the device is judged to be in a state of adjustment lag and execution unit response failure.

[0055] A fault identification system for HVAC equipment includes a data acquisition module, a behavior-driven modeling module, an anomaly detection module, a fault correlation analysis module, and an early warning feedback module;

[0056] The data acquisition module is used to collect building usage behavior, environmental response data, and HVAC equipment operation data.

[0057] The behavior-driven modeling module is used to construct a coupled model and behavior-driven operation mode between building behavior and environment and equipment response.

[0058] The anomaly detection module is used to identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and to generate structured anomaly codes.

[0059] The fault correlation analysis module is used to correlate abnormal trajectories with equipment maintenance data, identify potential fault types, and output fault information.

[0060] The early warning feedback module is used to generate structured early warning data, visualize and display it, and track operation and maintenance responses.

[0061] The beneficial effects of this invention are as follows: This invention can integrate building behavior, environmental response and equipment operation data for modeling and analysis, improve the accuracy and real-time performance of fault identification, accurately identify abnormal behavior trajectories and response misalignments, assist in locating potential causes of equipment failures, achieve rapid feedback of fault information and operation and maintenance response through structured early warning and visualization, introduce causal correlation analysis, effectively utilize historical maintenance data, and improve the intelligence of fault judgment and maintenance efficiency. Attached Figure Description

[0062] Figure 1 A flowchart illustrating the steps of a fault identification method for HVAC equipment according to an embodiment of the present invention;

[0063] Figure 2 This is a basic flowchart of a fault identification system for HVAC equipment provided in one embodiment of the present invention. Detailed Implementation

[0064] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0065] Example 1, referring to Figure 1 A method for identifying faults in HVAC equipment is provided, comprising the following steps:

[0066] Step S1: Collect building usage behavior, environmental response data, and HVAC equipment operation data.

[0067] Step S2: Construct a coupling model and behavior-driven operation mode between building behavior and environment and equipment response.

[0068] Step S3: Identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and generate structured anomaly codes.

[0069] Step S4: Associate the abnormal trajectory with equipment maintenance data, identify potential fault types, and output fault information.

[0070] Step S5: Generate structured early warning data, visualize and track operation and maintenance responses.

[0071] This invention can integrate building behavior, environmental response, and equipment operation data for modeling and analysis, improving the accuracy and real-time performance of fault identification. It can accurately identify abnormal behavior trajectories and response misalignments, assisting in locating potential causes of equipment failures. Through structured early warning and visualization, it enables rapid feedback of fault information and operation and maintenance response. By introducing causal correlation analysis and effectively utilizing historical maintenance data, it improves the intelligence of fault judgment and maintenance efficiency.

[0072] Step S1 includes the following sub-steps:

[0073] Step S101: Collect building usage behavior data, which includes the frequency of personnel movement, the density of personnel movement, the passage time, the number of people passing through, the on / off status of the lighting system, and the duration of the lighting system being on.

[0074] Step S101 collects building usage behavior data such as the frequency and density of personnel movement, passage time, and lighting status, which can reflect the actual usage pattern of the building space and provide behavioral driving basis for analyzing environmental changes and equipment responses.

[0075] Step S102: Collect environmental response data, including ambient temperature, ambient relative humidity, and ambient carbon dioxide concentration.

[0076] Step S102 collects environmental response data such as ambient temperature, relative humidity, and carbon dioxide concentration, which can characterize the indoor environmental status in real time and provide a reference for judging the relationship between the equipment adjustment effect and environmental feedback.

[0077] Step S103: Collect HVAC equipment operation data, including equipment start-up and shutdown time, equipment start-up and shutdown frequency, fan operating speed, temperature difference between supply air outlet and return air outlet, and energy consumption data.

[0078] Step S103 collects data on equipment start-up and shutdown time, frequency, fan speed, supply and return air temperature difference, and energy consumption, which can comprehensively reflect the operating status of HVAC equipment and lay a data foundation for identifying equipment performance changes and abnormal responses.

[0079] Step S1 collects data on building usage behavior, environmental response, and HVAC equipment operation to form a multi-dimensional raw data foundation, providing comprehensive and accurate data support for subsequent dynamic coupling modeling and fault identification.

[0080] Step S2 includes the following sub-steps:

[0081] Step S201: Establish the time-series correspondence between building usage behavior data and environmental response data. The logic for establishing the time-series correspondence is as follows:

[0082] The building usage behavior data and environmental response data are time-aligned, and the building usage behavior data is converted into a standardized time series feature vector. Using minutes as the time window unit, a sliding window comparison process is performed to compare the data with the environmental response data within the same time window.

[0083] Calculate the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data within each time window. Record the trend changes of the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data over continuous time periods. Based on the correlation coefficient, cross-lag time difference, co-fluctuation amplitude, and trend changes between the rate of change of building use behavior data and the rate of change of environmental response data, establish a dynamic coupling model between building use behavior data and environmental response data, and extract the driving features of the dynamic coupling model, including the maximum correlation time lag, the difference in fluctuation amplitude, and the peak response duration.

[0084] Step S201 establishes a dynamic coupling model and extracts driving features by performing time alignment and sliding window analysis on building usage behavior data and environmental response data. This can quantify the impact of building behavior on environmental conditions and provide behavioral driving basis for subsequent equipment response analysis.

[0085] Step S202: Establish a regulation response mapping between environmental response data and HVAC equipment operating data. The logic for establishing the regulation response mapping is as follows:

[0086] Environmental response data is converted into a standardized time series and time-aligned with HVAC equipment operation data. Using minutes as the time window unit, a sliding window comparison is performed to analyze the correspondence between parameter fluctuations in environmental response data and changes in the data state of HVAC equipment operation data. The equipment response delay, adjustment amplitude, and response rate within each time window are calculated, and response features are extracted. The response features include adjustment lag time, adjustment intensity change amplitude, and duration of adjustment.

[0087] Step S202 involves analyzing the adjustment response mapping between environmental response data and HVAC equipment operation data to extract features such as equipment response delay, adjustment range, and duration, thereby achieving a precise correlation between equipment operating status and environmental feedback.

[0088] Step S203: Integrate building usage behavior data, environmental response data, and equipment operation data, using driving characteristics and response characteristics as joint inputs to construct a building behavior-driven operation mode. The construction logic of the building behavior-driven operation mode is as follows:

[0089] Using a sliding time window as the analysis unit, the collaborative change segments of building usage behavior data, environmental response data, and equipment operation data on the time axis are identified respectively. A multivariate dynamic regression model and an adaptive weight fusion mechanism are constructed to establish a behavior-driven equipment response prediction sub-model.

[0090] By using response features as mediating variables, path analysis is introduced to quantify the moderating effect of response features on equipment operation response driven by changes in building behavior. The moderating effect indicators include the direction, intensity, and hysteresis magnitude of the modulation. A linkage map is constructed, and the path intensity and response probability are dynamically updated and visualized using a data-driven approach, outputting a structured building behavior response feature vector.

[0091] Step S203 integrates behavior-driven features and equipment response features to construct a building behavior-driven operation mode, and outputs a structured building behavior response feature vector through path analysis and dynamic regression mechanism to provide comprehensive feature support for anomaly detection and fault determination.

[0092] Step S2 establishes a dynamic coupling relationship between building usage behavior, environmental response, and HVAC equipment operation data to achieve temporal correlation of multidimensional data and construction of behavior-driven operation modes, providing high-precision feature input for subsequent anomaly detection and fault identification.

[0093] Step S3 includes the following sub-steps:

[0094] Step S301 involves matching and analyzing the structured building behavior response feature vectors between building behavior and equipment operation. Based on the constructed dynamic mapping model, a one-to-one correspondence analysis is performed between the structured building behavior response feature vectors and the equipment operation state sequence within a unit time window. The synchronization rate is used to evaluate the temporal consistency.

[0095] If a misalignment, decrease in synchronization rate, or change in correlation coefficient from positive to negative is detected in the structured building behavior response feature vector within a continuous time window, it is considered an area of ​​inconsistent behavior response, and the corresponding spatial location identifier, start time, and duration are recorded.

[0096] Step S301 involves matching the structured building behavior response feature vector with the equipment operating status, calculating the synchronization rate and correlation coefficient, identifying the time-series inconsistency area between behavior-driven and equipment response, and recording the abnormal location and duration.

[0097] Step S302: Identify abnormal behavioral trajectories of equipment energy consumption curve deviation and response lag under specific behavioral patterns. Perform cluster analysis based on the structured building behavior response feature vector, output the behavior pattern clustering results, and select typical behavior cycles based on the behavior pattern clustering results. Typical behavior cycles include high-density office work, nighttime idleness, and concentrated meetings. Perform statistical analysis on the equipment energy consumption curve and operating parameters within each typical behavior cycle. The logic of the statistical analysis is as follows:

[0098] Compare the equipment energy consumption curve and operating parameters with the baseline operating template. If the equipment energy consumption curve shows a sudden increase, continuous high fluctuation, or significant response delay when the specific behavioral input remains unchanged, it is determined to be an abnormal trajectory.

[0099] By setting preset energy consumption offset thresholds and response time limits, the start and end times, behavior pattern labels, and fluctuation amplitudes of abnormal trajectories are calibrated, and abnormal behavior trajectories are output.

[0100] Abnormal behavior trajectories include one or more abnormal manifestations, which are identified and confirmed by combining the time period of occurrence, corresponding behavioral patterns, and operational status:

[0101] If, during periods of stable personnel flow and limited changes in heat load, the equipment energy consumption data curve shows non-periodic fluctuations and a sudden increase in energy consumption without obvious external driving conditions, the equipment is judged to be operating ineffectively and its energy consumption regulation has failed.

[0102] The equipment frequently starts and stops without corresponding changes in building usage data, indicating a problem with the equipment's start / stop logic combined with excessively high control sensitivity.

[0103] When environmental response data fluctuates rapidly, if the operating status data fails to respond in a timely manner or the delay exceeds the preset adjustment response threshold, the device is judged to be in a state of adjustment lag and execution unit response failure.

[0104] Step S302 identifies typical behavior cycles through cluster analysis, compares the energy consumption curve with the benchmark template, detects sudden increases in energy consumption, continuous high-level fluctuations and response lags, outputs abnormal behavior trajectories and determines abnormal types such as invalid operation, abnormal start-stop logic and adjustment lags.

[0105] Step S303: Jointly structure and encode the inconsistent response area and the abnormal behavior trajectory. The coding fields of the structure coding include the anomaly type, the scope of impact, the associated device number, the start and end timestamps, and the anomaly severity score.

[0106] Step S303 involves jointly encoding the inconsistent behavioral response area and the abnormal behavioral trajectory to form structured data containing the anomaly type, scope of impact, equipment number, and anomaly score, providing standardized information for subsequent fault association and early warning output.

[0107] Step S3 identifies temporal consistency anomalies and energy consumption deviations by matching building behavior response characteristics with equipment operating status, and generates structured anomaly codes, providing a high-precision anomaly data foundation for fault location and subsequent causal analysis.

[0108] Step S4 includes the following sub-steps:

[0109] Step S401: Obtain historical maintenance data and key component operating status data of the equipment, and retrieve maintenance records from the equipment management system. The maintenance records include key component replacement time, maintenance frequency, maintenance item details, fault description, handling method and actual handling time.

[0110] The system synchronously collects operational status data of key components, including fans, electric valves, sensors, and control actuators. The operational status data includes cumulative runtime, temperature rise curve slope, vibration spectrum peak value, and abnormal fluctuation amplitude of operating current.

[0111] Step S401 retrieves historical maintenance records of the equipment and synchronously collects the operating status of key components to form a complete equipment health data chain, providing basic information for causal analysis of abnormal trajectories and equipment status.

[0112] Step S402: Perform causal correlation analysis between the abnormal behavior trajectory and historical maintenance data. Using the associated device number and abnormal time period in the abnormal behavior trajectory output in step S303 as anchor points, retrieve the maintenance records and status fluctuation information of the corresponding device in the previous and subsequent cycles in the maintenance database.

[0113] A causal mapping rule table is constructed based on the equipment type and the corresponding component characteristics. It is determined whether there are any abnormal situations before the occurrence of abnormal trajectories. Abnormal situations include maintenance delay records, frequent failures, or significant deterioration trends in operating status indicators. Causal association is determined based on the abnormal situations, and causal associations are output. Causal associations include abnormal behavior and component aging, lack of lubrication, and transmission imbalance.

[0114] Causal association analysis specifically includes:

[0115] Based on the device status record data, compare the time when the adjustment signal is issued with the actual response time of the device, calculate the average response time, response variance and maximum delay time, and determine whether it exceeds the preset threshold.

[0116] The number of times equipment starts and stops within the time period of abnormal behavior trajectory is statistically analyzed, and combined with building usage behavior data, it is determined whether there are invalid responses with frequent starts and stops but no effective environmental regulation effect.

[0117] Pattern recognition is performed on the energy consumption curves within a time period. The current energy consumption curve is normalized and aligned with the historical benchmark energy consumption curves under similar operating conditions, and the deviation index is calculated. Combined with equipment aging label data, a weighted comprehensive scoring model is used to calculate the risk level and output the judgment result.

[0118] Step S402 performs causal correlation analysis between abnormal behavior trajectories and maintenance data. By combining indicators such as response delay, start-stop frequency, and energy consumption deviation with the causal mapping rule table, it determines whether the abnormality originates from component aging, lack of lubrication, or transmission imbalance, and outputs the causal correlation results and risk level.

[0119] Step S403: Based on abnormal conditions, energy consumption curve deviations and changes in the status of key components, determine whether there are hidden faults in the equipment. Hidden faults include air duct blockage, fan aging, sensor accuracy degradation and control logic delay.

[0120] The final fault type identification result is output by combining the causal mapping rule table, along with equipment identification, fault classification code, impact level score and suggested handling measures.

[0121] Step S403, based on causal analysis and the status of key components, identifies hidden faults such as duct blockage, fan aging, and sensor accuracy degradation, and outputs the final fault type, impact level, and recommended handling measures, providing accurate reference for equipment maintenance optimization.

[0122] Step S4 constructs causal mapping rules by associating abnormal behavior trajectories with historical maintenance and key component status data, achieving accurate matching between abnormal trajectories and hidden equipment faults, and outputting fault identification results with risk levels and handling suggestions, providing intelligent decision-making basis for equipment maintenance.

[0123] Step S5 includes the following sub-steps:

[0124] Step S501: Convert the final fault type identification result into structured early warning data, specifically including:

[0125] The system reads the device identifier, fault type, risk level, and recommended measures, and combines this with the spatial location coding information in the abnormal behavior trajectory to generate a standard format early warning data structure. The early warning data structure includes spatial location information, abnormal situation, hidden fault, time period, and device coverage area.

[0126] Step S501 converts the fault identification results into a standardized early warning data structure, which includes spatial location, anomaly type and risk level, to achieve structured storage and rapid retrieval of information.

[0127] Step S502: Visualize the structured early warning data and synchronize it to the operation and maintenance platform. Use layer overlay to highlight the identified abnormal areas on the building floor plan and display information cards in the corresponding areas. The information cards include the equipment number, abnormal situation, hidden fault, risk level score and recommended handling solution.

[0128] Step S502 highlights the abnormal area on the building floor plan through visualization and synchronizes the information card on the operation and maintenance platform, enabling operation and maintenance personnel to quickly locate faulty equipment and obtain handling suggestions.

[0129] Step S503: Update the dynamic mapping model based on the actual response situation and track the operation and maintenance response records after the early warning information is released. The actual response situation includes the response start time, handling method, handling duration and final fault confirmation result.

[0130] The actual response is compared with the structured early warning data to determine whether there are false alarms, missed alarms, and classification biases, and to correct the judgment thresholds and causal reasoning path parameters.

[0131] Step S503 compares the early warning data with the operation and maintenance response records, assesses the false alarms and missed alarms, and dynamically adjusts the judgment threshold and inference path to achieve continuous optimization and accuracy improvement of the fault identification model.

[0132] Example 2, refer to Figure 2 This paper provides a fault identification system for HVAC equipment, including a data acquisition module, a behavior-driven modeling module, an anomaly detection module, a fault correlation analysis module, and an early warning feedback module.

[0133] The data acquisition module is used to collect building usage behavior, environmental response data, and HVAC equipment operation data.

[0134] The behavior-driven modeling module is used to build a coupled model and behavior-driven operation mode between building behavior and environment and equipment response.

[0135] The anomaly detection module is used to identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and to generate structured anomaly codes.

[0136] The fault correlation analysis module is used to correlate abnormal trajectories with equipment maintenance data, identify potential fault types, and output fault information.

[0137] The early warning feedback module is used to generate structured early warning data, visualize and display it, and track operation and maintenance responses.

[0138] This invention improves the accuracy of equipment fault identification through multi-dimensional data acquisition and dynamic coupling modeling. It can monitor the correlation between equipment operating status and environmental changes in real time. By identifying abnormal patterns between building usage behavior and equipment response, it can accurately locate the potential causes of equipment faults, reducing misjudgments and omissions in traditional fault diagnosis. By adopting a structured early warning mechanism combined with visualization, it can promptly provide fault information and support rapid response from maintenance personnel, thus optimizing equipment maintenance cycles and management efficiency. The introduction of causal correlation analysis in equipment maintenance and fault judgment helps to extract valuable information from historical maintenance records, achieving intelligent and precise equipment maintenance.

[0139] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0140] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for identifying faults in HVAC equipment, characterized in that, Includes the following steps: Step S1: Collect building usage behavior, environmental response data, and HVAC equipment operation data; Step S2: Construct a coupling model and behavior-driven operation mode between building behavior and environment and equipment response; Step S3: Identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and generate structured anomaly codes; Step S4: Associate the abnormal trajectory with equipment maintenance data, identify potential fault types, and output fault information; Step S5: Generate structured early warning data, visualize and track operation and maintenance responses; Step S2 includes the following sub-steps: Step S201: Establish a temporal correspondence between building usage behavior data and environmental response data. The logic for establishing the temporal correspondence is as follows: The building usage behavior data and environmental response data are time-aligned, the building usage behavior data is converted into a standardized time series feature vector, and a sliding window comparison process is performed with minutes as the time window unit, comparing it with the environmental response data within the same time window. Calculate the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data within each time window. Record the trend changes of the correlation coefficient, cross-lag time difference, and co-fluctuation amplitude between the rate of change of building use behavior data and the rate of change of environmental response data over a continuous period of time. Based on the correlation coefficient, cross-lag time difference, co-fluctuation amplitude, and trend changes between the rate of change of building use behavior data and the rate of change of environmental response data, establish a dynamic coupling model between building use behavior data and environmental response data, and extract the driving features of the dynamic coupling model. The driving features include the maximum correlation time lag, the difference in fluctuation amplitude, and the peak response duration. Step S202: Establish a regulation response mapping between environmental response data and HVAC equipment operation data. The logic for establishing the regulation response mapping is as follows: The environmental response data is converted into a standardized time series and time-aligned with the HVAC equipment operation data. Using minutes as the time window unit, a sliding window comparison is performed to analyze the correspondence between the parameter fluctuations of the environmental response data and the data state changes of the HVAC equipment operation data. The equipment response delay, adjustment amplitude, and response rate within each time window are calculated, and response features are extracted. The response features include adjustment lag time, adjustment intensity change amplitude, and continuous adjustment time. Step S203: Integrate building usage behavior data, environmental response data, and equipment operation data, and use driving features and response features as joint inputs to construct a building behavior-driven operation mode. The construction logic of the building behavior-driven operation mode is as follows: Using a sliding time window as the analysis unit, the collaborative change segments of building usage behavior data, environmental response data and equipment operation data on the time axis are identified respectively. A multivariate dynamic regression model and an adaptive weight fusion mechanism are constructed to establish a behavior-driven equipment response prediction sub-model. By using response features as mediating variables, path analysis is introduced to quantify the moderating effect of response features on equipment operation response driven by changes in building behavior. The moderating effect index includes the direction of adjustment, intensity, and hysteresis amplitude. A linkage map is constructed, and the path intensity and response probability are dynamically updated and visualized using a data-driven approach, outputting a structured building behavior response feature vector.

2. The HVAC equipment fault identification method as described in claim 1, characterized in that, Step S1 includes the following sub-steps: Step S101: Collect building usage behavior data, which includes the frequency of personnel movement, the density of personnel movement, the passage time, the number of people passing through, the on / off status of the lighting system, and the duration of the lighting system being on. Step S102: Collect environmental response data, including ambient temperature, ambient relative humidity, and ambient carbon dioxide concentration; Step S103: Collect HVAC equipment operation data, which includes equipment start-up and shutdown time, equipment start-up and shutdown frequency, fan operating speed, temperature difference between supply air outlet and return air outlet, and energy consumption data.

3. The HVAC equipment fault identification method as described in claim 2, characterized in that, Step S3 includes the following sub-steps: Step S301 involves matching and analyzing the structured building behavior response feature vectors between building behavior and equipment operation. Based on the constructed dynamic mapping model, a one-to-one correspondence analysis is performed between the structured building behavior response feature vectors and the equipment operation state sequence within a unit time window. The synchronization rate is used to evaluate the temporal consistency. If a misalignment, decrease in synchronization rate, or change in correlation coefficient from positive to negative is detected in the structured building behavior response feature vector within a continuous time window, it is considered an area of ​​inconsistent behavior response, and the corresponding spatial location identifier, start time, and duration are recorded.

4. The HVAC equipment fault identification method as described in claim 3, characterized in that, Step S3 further includes the following sub-steps: Step S302: Identify abnormal behavioral trajectories of equipment energy consumption curve deviation and response lag under specific behavioral patterns. Perform cluster analysis based on the structured building behavior response feature vector, output behavioral pattern clustering results, and select typical behavioral cycles based on the behavioral pattern clustering results. The typical behavioral cycles include high-density office work, nighttime idleness, and concentrated meetings. Perform statistical analysis on the equipment energy consumption curve and operating parameters within each typical behavioral cycle. The logic of the statistical analysis is as follows: Compare the equipment energy consumption curve and operating parameters with the baseline operating template. If the equipment energy consumption curve shows a sudden increase, continuous high fluctuation, or significant response delay under the condition that the specific behavioral input remains unchanged, it is judged as an abnormal trajectory. By setting preset energy consumption offset threshold and response time upper limit, the start and end time, behavior pattern label and fluctuation amplitude of abnormal trajectory are calibrated, and abnormal behavior trajectory is output. Step S303: Jointly structure and encode the inconsistent response area and the abnormal behavior trajectory. The encoded fields of the structured encoding include the abnormality type, the scope of impact, the associated device number, the start and end timestamps, and the abnormality score.

5. The HVAC equipment fault identification method as described in claim 4, characterized in that, Step S4 includes the following sub-steps: Step S401: Obtain historical maintenance data and key component operating status data of the equipment, and retrieve maintenance records from the equipment management system. The maintenance records include key component replacement time, maintenance frequency, maintenance item details, fault description, handling method and actual handling time. The system synchronously collects operational status data of key components, including fans, electric valves, sensors, and control execution units. The operational status data includes cumulative running time, temperature rise curve slope, vibration spectrum peak value, and abnormal fluctuation amplitude of operating current. Step S402: Perform causal correlation analysis between the abnormal behavior trajectory and historical maintenance data. Using the associated device number and abnormal time period in the abnormal behavior trajectory output in step S303 as anchor points, retrieve the maintenance records and status fluctuation information of the corresponding device in the previous and subsequent cycles in the maintenance database. A causal mapping rule table is constructed based on the equipment type and the corresponding component characteristics. It is determined whether there are any abnormal situations before the occurrence of abnormal trajectories. The abnormal situations include maintenance delay records, frequent failures, or significant deterioration trends in operating status indicators. Causal association is determined based on the abnormal situations, and the causal association is output. The causal association includes abnormal behavior and component aging, lack of lubrication, and transmission imbalance. Step S403: Based on abnormal conditions, energy consumption curve deviations and changes in the status of key components, determine whether there are hidden faults in the equipment. The hidden faults include air duct blockage, fan aging, sensor accuracy degradation and control logic delay. The final fault type identification result is output by combining the causal mapping rule table, along with equipment identification, fault classification code, impact level score and suggested handling measures.

6. The HVAC equipment fault identification method as described in claim 5, characterized in that, Step S5 includes the following sub-steps: Step S501: Convert the final fault type identification result into structured early warning data, specifically including: Read the device identifier, fault type, risk level and recommended measures, and combine them with the spatial location coding information in the abnormal behavior trajectory to generate a standard format early warning data structure. The early warning data structure includes spatial location information, abnormal situation, hidden fault, time period and device coverage area. Step S502: Visualize the structured early warning data and synchronize it to the operation and maintenance platform. Use layer overlay to highlight the identified abnormal areas on the building floor plan and display information cards in the corresponding areas. The information cards include the equipment number, abnormal situation, hidden fault, risk level score and recommended handling solution. Step S503: Update the dynamic mapping model based on the actual response situation and track the operation and maintenance response records after the early warning information is released. The actual response situation includes the response start time, handling method, handling duration and final fault confirmation result. The actual response is compared with the structured early warning data to determine whether there are false alarms, missed alarms, and classification biases, and to correct the judgment thresholds and causal reasoning path parameters.

7. The HVAC equipment fault identification method as described in claim 6, characterized in that, The causal relationship analysis specifically includes: Based on the device status record data, compare the time when the adjustment signal is issued with the actual response time of the device, calculate the average response time, response variance and maximum delay time, and determine whether it exceeds the preset threshold. The number of times the equipment starts and stops within the time period of abnormal behavior trajectory is statistically analyzed, and combined with building usage behavior data, it is determined whether there are invalid responses with frequent starts and stops but no effective environmental regulation effect; Pattern recognition is performed on the energy consumption curves within the specified time period. The current energy consumption curve is normalized and aligned with the historical benchmark energy consumption curves under similar operating conditions, and the deviation index is calculated. Combined with equipment aging label data, a weighted comprehensive scoring model is used to calculate the risk level and output the judgment result.

8. The HVAC equipment fault identification method as described in claim 7, characterized in that, The abnormal behavior trajectory includes one or more abnormal manifestations, which are identified and confirmed by combining the time period of occurrence, corresponding behavior pattern, and operating status: During periods of stable personnel flow and limited changes in heat load, if the equipment energy consumption data curve shows non-periodic fluctuations and a sudden increase in energy consumption without obvious external driving conditions, the equipment is judged to be ineffective in operation and energy consumption regulation failure. The equipment frequently starts and stops without corresponding changes in building usage data, indicating that the problem stems from an abnormal start / stop logic combined with excessively high control sensitivity. When environmental response data fluctuates rapidly, if the operating status data fails to respond in a timely manner or the delay exceeds the preset adjustment response threshold, the device is judged to be in a state of adjustment lag and execution unit response failure.

9. A fault identification system for heating, ventilation, and air conditioning (HVAC) equipment, applied in a fault identification method for HVAC equipment as described in any one of claims 1-8, characterized in that, It includes a data acquisition module, a behavior-driven modeling module, an anomaly detection module, a fault correlation analysis module, and an early warning feedback module; The data acquisition module is used to collect building usage behavior, environmental response data, and HVAC equipment operation data. The behavior-driven modeling module is used to construct a coupled model and behavior-driven operation mode between building behavior and environment and equipment response. The anomaly detection module is used to identify areas of inconsistent behavioral responses and abnormal behavioral trajectories, and to generate structured anomaly codes. The fault correlation analysis module is used to correlate abnormal trajectories with equipment maintenance data, identify potential fault types, and output fault information. The early warning feedback module is used to generate structured early warning data, visualize and display it, and track operation and maintenance responses.

Images